{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/word2vec-applied-to-recommendation","title":"Word2Vec applied to Recommendation: Hyperparameters Matter","arxiv_id":"1804.04212","date":"2018-04-11","proceeding":null,"authors":["Hugo Caselles-Dupré","Florian Lesaint","Jimena Royo-Letelier"],"abstract":"Skip-gram with negative sampling, a popular variant of Word2vec originally\ndesigned and tuned to create word embeddings for Natural Language Processing,\nhas been used to create item embeddings with successful applications in\nrecommendation. While these fields do not share the same type of data, neither\nevaluate on the same tasks, recommendation applications tend to use the same\nalready tuned hyperparameters values, even if optimal hyperparameters values\nare often known to be data and task dependent. We thus investigate the marginal\nimportance of each hyperparameter in a recommendation setting through large\nhyperparameter grid searches on various datasets. Results reveal that\noptimizing neglected hyperparameters, namely negative sampling distribution,\nnumber of epochs, subsampling parameter and window-size, significantly improves\nperformance on a recommendation task, and can increase it by an order of\nmagnitude. Importantly, we find that optimal hyperparameters configurations for\nNatural Language Processing tasks and Recommendation tasks are noticeably\ndifferent.","url_abs":"http://arxiv.org/abs/1804.04212v3","url_pdf":"http://arxiv.org/pdf/1804.04212v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"word2vec-applied-to-recommendation","repo_url":"https://github.com/deezer/w2v_reco_hyperparameters_matter","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"word-embeddings","task_name":"Word Embeddings"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1804.04212","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}